Could you actually save money on solar PV?

There are several `home power' renewable energy techologies available to consumers today.  The big two are wind and solar.  Home solar systems include water heating, air heating, and solar photovoltaics (PV).  I intend only to address electrical generation here, so I will ignore solar heat and hot-water systems.

Wind power is substantially cheaper than solar, but in the interests of good relations with your neighbors, it is often unsuitable to suburban housing.  In particular, you have to put up a support tower to get the wind turbine into the wind flow, and the blades tend to make noise.  You might find objections to the `unsightly' or `noisy' device.  They might even be from your wife (or husband).  Solar PV, on the other hand, is at worst sometimes considered inappropriate for roof decor, e.g., in an area where the local homeowner's association requires everyone to have red clay tile roofs.  In areas with asphalt shingle roofing, or any kind of roofing mix, objections are rare.  All you need, then, is a suitable expanse of south- or west-facing roof that gets full sun for much of the day.

(If for some reason your neighborhood does object to rooftop PV, you may be able to use backyard PV instead.  You might consider building a `solar gazebo' or `solar ramada', for instance.  This will cut down on the available square footage though.  PV systems usually work out to about 10 watts per square foot, so for 1.5 kW, you need 150 sqft or so.)

Note that I am about to describe a large system; smaller ones are cheaper and will generally pay back better.  I wanted to use a more dramatic example though.

The main drawback to solar PV is that it is very expensive today.  A big system, such as the 3 kW one in the worksheet below, might easily cost $25,000 installed.  The advantage is that such a system has no moving parts and requires no maintenance, other than perhaps an occasional hose-down of the panels if they get dusty.  If you choose to include a battery backup in your system, the batteries will require maintenance.  This example assumes a pure grid-tied system, retrofitted to an existing home, with all the work being done by your system installer.

Of course, $25,000 is a lot of money -- you could buy a nice car for that price -- but one advantage you have here is that you can roll it into a home mortgage.  For instance, you could refinance your house to take advantage of today's lower mortgage rates, extract some equity in the process, and apply that to the PV system.  (Tell your lender that you are going to use it for `home efficiency improvements' so as not to scare them.  Seriously, if they ask, tell them the truth, but be aware that bankers may be dubious.  Solar PV, that's some newfangled thing isn't it?  Pardon me, I have to go feed the horses before tonight's trip home.  By the way, do you know a good place to buy a new buggy-whip?  I keep hearing about these `automobiles' but I don't trust them yet either.)

Earlier this year, I made up some spreadsheets to figure out when and where solar PV might be economical.  In the spreadsheet below, I used insolation (`average sunshine equivalent', more or less) figures I found for Riverside, and put in the Southern California Edison baseline figures for a gas-heated house in the territory that includes the Riverside area.  I then set the average summer use at about 1000 kWh per month, and the average winter use somewhat lower.

Note that SCE considers `summer' to be only four months long, rather than six.  For this reason, the insolation figure for winter below probably underestimates the power output (and hence the dollar savings) -- my spreadsheet uses the `winter sun' figure for all 8 months of the `winter baseline', when in fact, the winter sun figure is really the minimum, i.e., what you get in December when the days are shortest.  (The National Renewable Energy Laboratory figures are set up this way so that someone building an off-grid house knows how much power they can count on to charge batteries in mid-winter.)  For Riverside, winter and summer are not all that different, so it may not really matter too much after all.

The numbers I used as inputs are marked `<' below.  This system reduces summer bills by more than $125 each month.  The average monthly savings comes to $110.38.  You might take the system cost of $25,500 -- which, by the way, is just a rough estimate that I got from my own local installer; it could vary for other areas -- and divide by $110.38 to get a payback time of 231 months, or about 19 1/4 years.  (Note that solar panels usually have a 20 or 25 year warranty and the overall system should work fine for at least 30 years, if not longer.  The least reliable part of a typical solar power system is the batteries.  The inverters also sometimes fail; they usually have a 5 year warranty.  The panels never seem to fail, and in fact, in one installation in the Sacramento area, most of the panels survived a windstorm that completely wrecked the structure they were built on.  As I recall, it was some kind of carport, and it blew away in high winds.  They retrieved the panels and just re-installed them on the replacement carport.)

Saving $110 every month sounds pretty reasonable, until you remember that you have to pay interest on a mortgage.  The payment on a 30-year, $25,500 loan at 6 3/4% is $165.39/mo.  So if you save, on average, $110.38, but have to pay $165.39, you will have a net loss of $55.01 each month.  Or will you?

Home mortgage interest is tax-deductible!  If you have a marginal tax rate of 36%, that $165.39/mo is really only $105.85 after taxes.  (Actually I did this too simply -- some of that is non-deductible principal.  Take these numbers with a grain of salt.)  Of course, this tax deduction eventually fades away as the interest is paid off, but for at least the first five years or so, you will have a net savings of more than $4.50/mo.  In other words, this solar PV setup would actually save you $54 each year.  What you are doing here is taking a tax deduction on your electric bill.

The keys to making this work are:

Tier 5, whose existence is rumored but not yet certain, appears in the spreadsheet below, but is not used in the example.  (The 1000 kWh/mo in summer, and 800 kWh/mo in winter, are within the 300% over baseline limits.)  Your goal, if you put in a PV system in order to save money, is to get rid of as much Tier 5 and 4 electricity as you can, and maybe some Tier 3 electricity too.  You do not need a really big 3 kW system to do this; a smaller system would actually have a better payback.  If you go too small, however, the fixed costs mount up -- my local installer quoted about $13,000 for 1 kW, versus about $18,000 for 2 kW.  If you really want to splurge, they will install 12 kW for $90,000 -- but 12 kW would generate about twice as much as the house below is shown as using, and the utility companies do not pay you anywhere near what this would be worth.  (In fact, PG&E does not pay you anything at all!)
Note: there is a way to make a solar PV system pay off even faster, using `net metering' and `time of use' metering.  Getting it all set up is something of a hassle, though.  I heard a story of one guy here in northern California who set up his solar system, paid PG&E to come out and put in a net-metering time-of-use meter, and then discovered that not only was he paying PG&E for electricity he used from them, he was also paying them -- instead of them paying him -- for electricity they used from him!  It took months to straighten it all out.  So, I have ignored that here, and just concentrated on what happens if you put in less solar PV than you actually use, so that your system just reduces your regular utility bill.

If you want to see whether time-of-use net metering might work for you, you will need to find your own utility's TOU rates.  For PG&E, it works out to about $0.33/kWh noon to 6 PM summer weekdays, and the trick is to make sure that you never use any PG&E power during that period, and sell them as much of your solar PV output as you can during that same period.  After 6 PM, your TOU meter charges you about $0.08/kWh.  That means each `peak summer kWh' you sell buys you 4 kWh at night.  In winter, the noon-to-6 cost is about $0.11/kWh, so you really only win in the summer.

I made two more spreadsheets for SoCal: super-cheap system (you do lots of footwork and if anything goes wrong you have to figure out what and hassle whichever vendor sold you bum hardware), and: medium-size system (local installer's ballpark figure for 1.8 kW system with full warranty etc.).  The medium-size system pays off about the same as the larger one below, and still gives you the ability to let someone else do all the work, but does not cost as much up front (about $16k instead of over $25k).

Back to Deregulation Fiasco

             (SoCalEdision variant)       

You must input the following:   
  your summer and winter baseline quantities        
  your summer and winter usage  
  your summer and winter `peak sun hours' (insolation)        

The baseline quantities can be found on your bills, and go by 
`territory' and whether you have electric heat.  Summer is defined      
as June through September, give or take a bit.  The baseline is         
given in kWh/day.  For instance baseline for SCE region 17 is 
13.1/10.5 (summer/winter) or 16.9/24.1 (all-electric).  By contrast,    
region 10 is 9.1/9.2 (summer/winter) or 10.0/16.2 (all-electric).       
(Interestingly, this means `summer' is 122 days and `winter' is         
243.25 days.)

Peak sun hours depend on location and local weather conditions.  I      
do not have much data on these but have sample cities.  Unfortunately   
the average listed in this data is not the computed average,  
which may throw some of these calculations off.     

  city             summer winter   average   average
                  (table)(table)   (table)   (s+w)/2
  Davis              6.09   3.31      5.10      4.70
  Fresno             6.19   3.42      5.38      4.80
  Inyokern           8.70   6.97      7.66      7.83
  La Jolla           5.24   4.29      4.77      4.77
  Los Angeles        6.14   5.03      5.62      5.58
  Riverside          6.35   5.35      5.87      5.85

You will also need a target system size (in watts or kW).  This is      
determined by the number of solar PV modules and their rated (peak)     
output.  Your inverter will have to be hefty enough for this as         
well.  The system size will affect the price, which you must also       
enter.  Finally you have to enter a system efficiency factor, to        
account for losses in the modules and inverter.  Efficiencies of        
82, 71, and 75 percent are typical for batteryless or battery-
backed-up systems.  The 75% efficient battery systems are slightly      
more expensive.          

Finally, you need to enter your utility's rates and tiers, any
local city taxes you pay on electricity, and -- for financial 
comparison purposes -- the rate you would expect to pay on a 30         
year mortgage, and your marginal tax rate for tax deductability.        


                    30.50 summer days/mo           4 smmr mos 
                    30.41 winter days/mo           8 wntr mos 

             UTILITY RATES      

         tier        rate  limit  smmr kWh  wntr kWh
          I        $.1301   100%    399.55    319.27
          II       $.1516   130%    119.87     95.78
          III      $.2300   200%    279.68    223.49
          IV       $.2800   300%    399.55    319.27
          V        $.4000   none      rest      rest


     baseline       13.10< summer (kWh/day)         
  (region 17)       10.50< winter (kWh/day)         

        usage       33.00< summer (kWh/day)         
                    26.00< winter (kWh/day)         

   insolation        6.35< summer (PSH/day)         
  (sun hours)        5.35< winter (PSH/day)         

  system size        3.00< kW   
  system cost  $25,500.00<      
   efficiency         82%<      

     city tax        5.0%<      

    mtge rate       6.75%<           36.0%< your tax bracket  
         term          30< years

             RESULTS - SUMMARY  

       w/o PV     $202.18/smmrmo   $172.41 avg.     

         w/PV      $76.23/smmrmo    $62.03 avg.     

        saves     $125.95/smmrmo   $110.38 avg.     

     mortgage     $165.39/mo       $105.85 (after tax equivalent)       
  net savings       $4.53

             RESULTS - DETAIL   

             without PV system  
       summer     1006.50 kWh:  
          I        399.55 kWh       $51.98
          II       119.87 kWh       $18.17
          III      279.68 kWh       $64.33
          IV       207.40 kWh       $58.07
          V          0.00 kWh        $0.00
                         subtot:   $192.55
                            tax:     $9.63

       winter      790.56 kWh   
          I        319.27 kWh       $41.54
          II        95.78 kWh       $14.52
          III      223.49 kWh       $51.40
          IV       152.03 kWh       $42.57
          V          0.00 kWh        $0.00
                         subtot:   $150.03
                            tax:     $7.50

             with PV system     
   summer gen      476.44 kWh/mo
   winter gen      400.18 kWh/mo             5107.18 kWh/yr   

       summer      530.06 kWh:  
          I        399.55 kWh       $51.98
          II       119.87 kWh       $18.17
          III       10.64 kWh        $2.45
          IV         0.00 kWh        $0.00
          V          0.00 kWh        $0.00
                         subtot:    $72.60
                            tax:     $3.63

       winter      390.39 kWh   
          I        319.27 kWh       $41.54
          II        71.12 kWh       $10.78
          III        0.00 kWh        $0.00
          IV         0.00 kWh        $0.00
          V          0.00 kWh        $0.00
                         subtot:    $52.32
                            tax:     $2.62